From mboxrd@z Thu Jan 1 00:00:00 1970 From: Linus Walleij Subject: Re: [PATCH 2/2] net/smsc911x: Add regulator support Date: Wed, 26 Oct 2011 11:25:09 +0200 Message-ID: <4EA7D1F5.2030904@stericsson.com> References: <1319616356-6391-1-git-send-email-linus.walleij@stericsson.com> <20111026083838.GA9157@opensource.wolfsonmicro.com> Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1"; format=flowed Content-Transfer-Encoding: 7bit Cc: "netdev@vger.kernel.org" , Steve Glendinning , Mathieu Poirer , Robert MARKLUND , Linus Walleij To: Mark Brown Return-path: Received: from eu1sys200aog107.obsmtp.com ([207.126.144.123]:49241 "EHLO eu1sys200aog107.obsmtp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932510Ab1JZJfV (ORCPT ); Wed, 26 Oct 2011 05:35:21 -0400 In-Reply-To: <20111026083838.GA9157@opensource.wolfsonmicro.com> Sender: netdev-owner@vger.kernel.org List-ID: On 10/26/2011 10:38 AM, Mark Brown wrote: > On Wed, Oct 26, 2011 at 10:05:56AM +0200, Linus Walleij wrote: > > >> + /* enable/disable regulator for vddvario */ >> + if (pdata->regulator_vddvario) { >> > This has the same issue as last time - if you've got conditional code > like this in the body of the driver something is going wrong. Unless > the supply is genuinely optional and might not be physically present on > some systems the driver should fail if it can't get it. The regulator > API will stub itself out when not in use. > That solves the issue for platforms with no regulator support at all. Then we have platforms with regulator support, but no regulator for this hardware, because that one happens to be always-on in these systems. And they do not have CONFIG_REGULATOR_DUMMY either. So the driver probe will fail. How do we solve this? Shall we have CONFIG_SMC911X select REGULATOR_DUMMY? Yours, Linus Walleij